OpenAI Launches Health In ChatGPT To Handle User Medical Records
OpenAI on Thursday, July 23, 2026 launched Health in ChatGPT, a U.S. feature that lets users upload medical records and link Apple Health.[1]
Health in ChatGPT is rolling out to logged-in U.S. users age 18 and older across Free, Go, Plus and Pro tiers on the web and iOS.[1] The interface lets people upload lab results, medication lists and other records, and it can summarize files and compare past and current data.[1] OpenAI says those records will not be used to train its models or to target ads and that the feature includes extra encryption and confirmation safeguards.[1]
On January 7, 2026, OpenAI first introduced ChatGPT Health, saying more than 230 million people were already asking health and wellness questions on ChatGPT each week. That initial launch gave limited access to testers on Free, Go, Plus and Pro plans outside the EEA, Switzerland and the UK while OpenAI refined models, including GPT-5.6 variants evaluated on health benchmarks. Tester feedback and model improvements prompted OpenAI's July 23 decision to broaden access to U.S. adults.
But experts warn of serious risks from AI-generated medical advice, citing legal cases including a Florida lawsuit over a pulmonary embolism and a California teen's death after following a ChatGPT-recommended substance mix.[1] Privacy specialists also caution that users are sharing deeply personal health information and question how durable OpenAI's privacy promises will be.[1] Some product team members and users say the tool helps people feel less lost in their health history, but others say accuracy, handoffs to clinicians and trust remain the biggest obstacles.
The mainstream summary does not address the regulatory landscape surrounding AI-generated medical advice, which is crucial given that consumer-facing large language models like ChatGPT Health are not classified as medical devices by the FDA, provided they include disclaimers about their intended use. This lack of regulation raises significant concerns about the reliability of the health guidance provided, as highlighted by the analysis from npj Digital Medicine. Furthermore, while the summary mentions user feedback indicating that the tool helps individuals feel less lost in their health history, it overlooks the broader context of dissatisfaction with the U.S. healthcare system. According to KFF's 2026 Health Tracking Poll, many adults are skipping necessary care due to high costs, which may drive them to seek alternative solutions like AI tools for health management. This context suggests that the launch of Health in ChatGPT is not just a technological advancement but also a response to systemic healthcare issues affecting millions of Americans.
Additionally, the summary downplays the ongoing legal scrutiny surrounding the feature, as evidenced by a lawsuit filed just before the launch that sought to block it. This legal backdrop underscores the risks associated with AI handling sensitive health data, which could affect user trust and the feature's long-term viability. The perspectives from social media also emphasize that success hinges not only on the AI's ability to explain medical information but also on effectively managing privacy and ensuring accurate guidance, which the mainstream account fails to capture.
Show source details & analysis (1 source)
📊 Relevant Data
Health in ChatGPT is rolling out exclusively to logged-in U.S. users aged 18 and older across Free, Go, Plus, and Pro plans on web and iOS.
Launching Health in ChatGPT — OpenAI
Consumer-facing large language models providing health guidance are not regulated by the FDA as medical devices when they include disclaimers that they are not intended for diagnosis, treatment, or clinical decision support.
Unregulated large language models produce medical advice — npj Digital Medicine (Nature)
📌 Key Facts
- On Thursday, July 23, 2026, OpenAI announced Health in ChatGPT, a dedicated space for health use inside ChatGPT.
- The feature allows users to upload medical records such as lab results and medications and to connect Apple Health so ChatGPT can summarize and compare past and current data.
- OpenAI says Health records will not be used to train its models or target ads and that the feature includes extra encryption and confirmation safeguards.
- Experts cite serious risks from AI-generated medical advice, referencing a Florida lawsuit over a pulmonary embolism case and a California teen’s death after following a ChatGPT-recommended substance mix.
- Privacy experts warn users are sharing deeply personal health information and question how durable OpenAI’s current privacy commitments will be.
📰 Source Timeline (1)
Follow how coverage of this story developed over time